Leiomyosarcomas are a group of highly heterogeneous tumors differing in locations, pathological features, response to therapy, and propensity for metastasis. Historically, leiomyosarcomas originating from the gastrointestinal (GI) tract do not respond to conventional chemotherapy, whereas, leiomyosarcomas arising from other organ systems are more likely to respond to doxorubicin-based therapies. This observation was refined by others and it became clear that patients with (gastrointestinal stromal tumor) GIST had not only poor response rates to standard investigational therapy, but also a worse overall outcome. Recent discoveries made by basic scientists, pathologists and clinical investigators has led to an understanding of the biological importance of c-kit in GISTs and the development of one of the most exciting examples of targeted therapy to date. The success of the c-kit tyrosine kinase inhibitor STI-571 (Gleevec) has caught the attention of the scientific community. Understanding the mechanisms of response and resistance to STI-571 will broaden our understanding of cancer biology and lead to strategic approaches in the treatment of other malignancies. The aims of this study are to identify molecular events at the level of gene expression that are associated with prognosis, metastasis, and the dramatic responses to STI-571 that have been observed with GIST but not leiomyosarcomas arising outside of the GI tract. We will couple gene-expression-profiling studies with high throughput confirmation and functional validation studies to gain understanding of the mechanisms underlying the important clinical phenotypes. We will test three related hypotheses. 1) There are key cellular pathway genes whose differential expressions are responsible for the different prognoses of leiomyosarcoma and GISTs; 2) There are key metastasis-regulating genes whose expressions are responsible for different metastasis potentials of leiomyosarcoma; 3) There are key genes that regulate response to STI-571 treatment in GIST but resistance to STI-571 in leiomyosarcomas arising outside the GI tract. In this proposal, we will test these 3 hypotheses with 3 specific aims taking advantage of high throughput transcriptome profiling technologies, high throughput tissue microarray and real-time polymerase chain reaction (PCR) validation, and in vitro gene perturbation functional studies. Aim 1: To generate transcriptome data using a microarray technique and to identify genes associated with extremity leiomyosarcomas (including retroperitoneal, vascular, uterine and extremity sites of origin) and GISTs. Aim 2: To identify genes that are involved in leiomyosarcoma and GIST metastasis and functionally test those genes in an in vitro invasion assay. Aim 3: To identify genes that regulate cellular response and resistance to the tyrosine kinase inhibitor STI-571, a drug for treating GISTs that have c-kit mutations and activation. Through this hypothesis-driven and discovery-driven program, we anticipate discovery of markers for clinical prediction as well as understanding of the molecular basis for those important clinical phenotypes.